This package depends on the RecoBTag/PerformanceMeasurements package. First follow the instructions at https://github.com/cms-btv-pog/RecoBTag-PerformanceMeasurements/tree/8_0_X and then add the TagVarExtractor package
git clone -b 8_0_X https://github.com/cms-btv-pog/RecoBTag-TagVarExtractor.git RecoBTag/TagVarExtractor
scram b -j8
cd RecoBTag/TagVarExtractor/test/
cmsRun tagvarextractor_cfg.py maxEvents=20000 reportEvery=1000 wantSummary=True
This will produce a ROOT file called JetTaggingVariables.root
that can be used for TMVA training.
To create and submit Condor jobs that will process all files stored in an EOS directory, you can
use the createAndSubmitJobs.py
script as in the following example
./createAndSubmitJobs.py -i /eos/uscms/store/user/rsyarif/RadionToHHTo4B_M-1500_TuneZ2star_8TeV-nm-madgraph/Summer12_DR53X-PU_S10_START53_V19-v1_BTagAnalyzerLite_5_3_X_v1.01/ \
-c tagvarextractor_cfg.py -o output_dir -n 60 -p maxEvents=-1,reportEvery=1000,wantSummary=True --fnal
Run
./createAndSubmitJobs.py --help
for more information on the available input parameters and their meaning.